(Revised Abstract) Description: How the nervous system extracts a coherent view of our complex and dynamic world is not well understood. Current understanding of the neural circuitry involved in visual object recognition is based in large part on physiological recordings made during passive animals looking at static images. Real world vision differs dramatically from these conditions, as both the observer and objects in the world are dynamic. Further, active vision is purposive; an observer's percepts can guide actions that may in turn alter the visual environment. The objective of the experiments described in this proposal is to begin to investigate the neural basis of active vision by recording the neural activity generated in extrastriate visual areas of the temporal lobe while monkeys observe, monitor, and interact with complex visual forms over temporally extended periods.The first specific aim will explore how knowing when something will occur prepares the perceptual system to determine what has occured. The second aim will test the hypothesis that neural activity in IT is maintained not only by the visible presence of the object, but also when that object disappears momentarily from view but in a context consistent with its continued physical presence. The final aim considers whether either overt action directed toward an object, or planned action for an upcoming event, directly affect the responses of IT neurons to visual stimuli. Existing neurophysiological theories of object processing propose that cells in the temporal cortex are essential for matching acquired visual snapshots to stored representations, but they leave open the question of how these brain areas contribute to continuous visual experience. By recording from cells whose putative role is to convey information about the detailed world, it is expected that these experiments can provide essential empirical data directly addressing this issue, and in doing so, will force the consideration of theories of dynamic neural processing as opposed to traditional "snapshot" models of visual perception.